/* * Copyright (c) 2006-2023, RT-Thread Development Team * * SPDX-License-Identifier: Apache-2.0 * * Change Logs: * Date Author Notes * 2022-11-14 WangXiaoyao the first version * 2023-08-17 Shell Add unmap_range for MAP_PRIVATE * Support MAP_FIXED in aspace_map(), and * Add better support of permission in mmap */ /** * @brief Virtual Address Space */ #include #define DBG_TAG "mm.aspace" #define DBG_LVL DBG_INFO #include #include "avl_adpt.h" #include "mm_private.h" #include #include #include #include #define ALIGNED(addr) (!((rt_size_t)(addr) & ARCH_PAGE_MASK)) static void *_find_free(rt_aspace_t aspace, void *prefer, rt_size_t req_size, void *limit_start, rt_size_t limit_size, mm_flag_t flags); static int _unmap_range_locked(rt_aspace_t aspace, void *addr, size_t length); struct rt_aspace rt_kernel_space; static int _init_lock(rt_aspace_t aspace) { int err; MM_PGTBL_LOCK_INIT(aspace); err = rt_mutex_init(&aspace->bst_lock, "aspace", RT_IPC_FLAG_FIFO); return err; } rt_err_t rt_aspace_init(rt_aspace_t aspace, void *start, rt_size_t length, void *pgtbl) { int err = RT_EOK; if (pgtbl) { aspace->page_table = pgtbl; aspace->start = start; aspace->size = length; aspace->private_object = RT_NULL; err = _aspace_bst_init(aspace); if (err == RT_EOK) { /** * It has the side effect that lock will be added to object * system management. So it must be paired with a detach once * the initialization return successfully. */ err = _init_lock(aspace); } } else { err = -RT_EINVAL; } return err; } rt_aspace_t rt_aspace_create(void *start, rt_size_t length, void *pgtbl) { rt_aspace_t aspace = NULL; int err; RT_ASSERT(length <= 0 - (rt_size_t)start); aspace = (rt_aspace_t)rt_malloc(sizeof(*aspace)); if (aspace) { rt_memset(aspace, 0, sizeof(*aspace)); err = rt_aspace_init(aspace, start, length, pgtbl); if (err != RT_EOK) { LOG_W("%s(%p, %lx, %p): failed with code %d\n", __func__, start, length, pgtbl, err); rt_free(aspace); aspace = RT_NULL; } } return aspace; } void rt_aspace_detach(rt_aspace_t aspace) { rt_varea_t varea; WR_LOCK(aspace); varea = ASPACE_VAREA_FIRST(aspace); while (varea) { rt_varea_t prev = varea; varea = ASPACE_VAREA_NEXT(varea); _varea_uninstall_locked(prev); if (VAREA_NOT_STATIC(prev)) { rt_free(prev); } } WR_UNLOCK(aspace); rt_aspace_anon_ref_dec(aspace->private_object); rt_mutex_detach(&aspace->bst_lock); } void rt_aspace_delete(rt_aspace_t aspace) { RT_ASSERT(aspace); rt_aspace_detach(aspace); rt_free(aspace); } rt_inline rt_size_t _get_effect_attr(rt_aspace_t aspace, rt_varea_t varea) { rt_size_t attr = varea->attr; /* not write permission for user on private mapping */ if (rt_varea_is_private_locked(varea)) attr = rt_hw_mmu_attr_rm_perm(attr, RT_HW_MMU_PROT_USER | RT_HW_MMU_PROT_WRITE); return attr; } static int _do_named_map(rt_aspace_t aspace, rt_varea_t varea, void *vaddr, rt_size_t length, rt_size_t offset, rt_size_t attr) { LOG_D("%s: va %p length %p", __func__, vaddr, length); int err = RT_EOK; /* it's ensured by caller that (void*)end will not overflow */ void *phyaddr = (void *)(offset << MM_PAGE_SHIFT); void *ret; attr = _get_effect_attr(aspace, varea); ret = rt_hw_mmu_map(aspace, vaddr, phyaddr, length, attr); if (ret == RT_NULL) { err = -RT_ERROR; } if (err == RT_EOK) rt_hw_tlb_invalidate_range(aspace, vaddr, length, ARCH_PAGE_SIZE); return err; } rt_inline void _do_page_fault(struct rt_aspace_fault_msg *msg, rt_size_t off, void *vaddr, rt_mem_obj_t mem_obj, rt_varea_t varea) { msg->off = off; msg->fault_vaddr = vaddr; msg->fault_op = MM_FAULT_OP_READ; msg->fault_type = MM_FAULT_TYPE_PAGE_FAULT; msg->response.status = MM_FAULT_STATUS_UNRECOVERABLE; msg->response.vaddr = 0; msg->response.size = 0; RT_ASSERT(mem_obj->on_page_fault); mem_obj->on_page_fault(varea, msg); } int rt_varea_map_with_msg(rt_varea_t varea, struct rt_aspace_fault_msg *msg) { rt_aspace_t aspace; int err = -RT_ERROR; if (msg->response.status == MM_FAULT_STATUS_OK) { /** * the page returned by handler is not checked * cause no much assumption can make on it */ char *store = msg->response.vaddr; rt_size_t store_sz = msg->response.size; if ((char *)msg->fault_vaddr + store_sz > (char *)varea->start + varea->size) { LOG_W("%s: too much (0x%lx) of buffer on vaddr %p is provided", __func__, store_sz, msg->fault_vaddr); } else { void *map; rt_size_t attr; void *v_addr = msg->fault_vaddr; void *p_addr = store + PV_OFFSET; aspace = varea->aspace; RT_ASSERT(aspace); attr = _get_effect_attr(aspace, varea); map = rt_hw_mmu_map(aspace, v_addr, p_addr, store_sz, attr); if (!map) { LOG_W("%s: MMU mapping failed for va %p to %p of %lx", __func__, msg->fault_vaddr, store + PV_OFFSET, store_sz); } else { rt_hw_tlb_invalidate_range(varea->aspace, v_addr, store_sz, ARCH_PAGE_SIZE); err = RT_EOK; } } } else if (msg->response.status == MM_FAULT_STATUS_OK_MAPPED) { if (rt_hw_mmu_v2p(varea->aspace, msg->fault_vaddr) == ARCH_MAP_FAILED) { LOG_D("%s: no page is mapped on %p", __func__, msg->fault_vaddr); } err = RT_EOK; } else { LOG_W("%s: failed on va %p inside varea %p(%s)", __func__, msg->fault_vaddr, varea, varea->mem_obj->get_name ? varea->mem_obj->get_name(varea) : "unknow"); } return err; } /* allocate memory page for mapping range */ static int _do_prefetch(rt_aspace_t aspace, rt_varea_t varea, void *start, rt_size_t size) { int err = RT_EOK; /* it's ensured by caller that start & size ara page-aligned */ char *end = (char *)start + size; char *vaddr = start; rt_size_t off = varea->offset + ((vaddr - (char *)varea->start) >> ARCH_PAGE_SHIFT); while (vaddr != end) { /* TODO try to map with huge TLB, when flag & HUGEPAGE */ struct rt_aspace_fault_msg msg; _do_page_fault(&msg, off, vaddr, varea->mem_obj, varea); if (rt_varea_map_with_msg(varea, &msg)) { err = -RT_ENOMEM; break; } /** * It's hard to identify the mapping pattern on a customized handler * So we terminate the prefetch process on that case */ if (msg.response.status == MM_FAULT_STATUS_OK_MAPPED) break; vaddr += msg.response.size; off += msg.response.size >> ARCH_PAGE_SHIFT; } return err; } static rt_err_t _expand_varea(rt_varea_t varea, void *new_va, rt_size_t size) { rt_err_t error; rt_aspace_t aspace; void *old_va; if (varea->mem_obj && varea->mem_obj->on_varea_expand) error = varea->mem_obj->on_varea_expand(varea, new_va, size); else error = -RT_EPERM; if (error == RT_EOK) { aspace = varea->aspace; old_va = varea->start; varea->size = size; if (old_va != new_va) { varea->start = new_va; varea->offset += ((long)new_va - (long)old_va) >> MM_PAGE_SHIFT; _aspace_bst_remove(aspace, varea); _aspace_bst_insert(aspace, varea); } } return error; } struct _mapping_property { rt_size_t attr; rt_size_t flags; rt_size_t offset; struct rt_mem_obj *mem_obj; }; #define INIT_PROP(obj,off,fl,attr) \ {.mem_obj = (obj), \ .offset = (off), \ .flags = (fl), \ .attr = (attr),} static rt_bool_t _contiguous_offset(rt_varea_t neighbour, rt_size_t map_size, struct _mapping_property *prop) { rt_size_t n_off = neighbour->offset; rt_size_t map_off = prop->offset; return n_off < map_off ? n_off + (neighbour->size >> MM_PAGE_SHIFT) == map_off : map_off + (map_size >> MM_PAGE_SHIFT) == n_off; } static rt_bool_t _compatible(rt_varea_t neighbour, rt_size_t map_size, struct _mapping_property *prop) { return (prop->attr == neighbour->attr && prop->flags == neighbour->flag && prop->mem_obj == neighbour->mem_obj && _contiguous_offset(neighbour, map_size, prop)); } rt_inline rt_err_t _migrate_and_release_varea(rt_aspace_t aspace, rt_varea_t to, rt_varea_t from, rt_err_t (*on_varea_merge)(struct rt_varea *to, struct rt_varea *from)) { rt_err_t error; error = on_varea_merge(to, from); if (error == RT_EOK) { /* uninstall operand & release the varea */ _aspace_bst_remove(aspace, from); to->size += from->size; if (VAREA_NOT_STATIC(from)) rt_free(from); } return error; } static rt_varea_t _merge_surrounding(rt_aspace_t aspace, rt_varea_t operand, struct _mapping_property *prop) { int again; rt_err_t error; int can_merge_fw; int can_merge_bw; rt_varea_t neighbour; char *operand_start; size_t operand_size; rt_err_t (*on_varea_merge)(struct rt_varea *to, struct rt_varea *from); if (operand->mem_obj && operand->mem_obj->on_varea_merge) { on_varea_merge = operand->mem_obj->on_varea_merge; do { operand_start = operand->start; operand_size = operand->size; LOG_D("search op_start=%p,op_size=0x%lx", operand_start, operand_size); /* find a compatible neighbour if any and setup direction */ can_merge_fw = can_merge_bw = 0; neighbour = _aspace_bst_search(aspace, operand_start - 1); if (!neighbour || !_compatible(neighbour, operand_size, prop)) { neighbour = _aspace_bst_search(aspace, operand_start + operand_size); if (neighbour && _compatible(neighbour, operand_size, prop)) can_merge_bw = 1; } else can_merge_fw = 1; if (can_merge_fw || can_merge_bw) { /* merge operand with its predecessor or successor */ if (can_merge_fw) { error = _migrate_and_release_varea(aspace, neighbour, operand, on_varea_merge); operand = neighbour; } else error = _migrate_and_release_varea(aspace, operand, neighbour, on_varea_merge); if (error == RT_EOK) again = 1; } else again = 0; } while (again); } return operand; } /** * Brief: expand and merge surrounding until not possible and * setup the pvarea if new virt address region is installed */ static rt_bool_t _try_expand_and_merge_okay(rt_aspace_t aspace, rt_varea_t *pvarea, void *alloc_va, rt_mm_va_hint_t hint, struct _mapping_property *prop) { int can_expand_fw; int can_expand_bw; rt_varea_t neighbour; rt_varea_t new_region_at = RT_NULL; rt_bool_t install_ok = RT_FALSE; char *operand_start = alloc_va; size_t operand_size = hint->map_size; /* find a compatible neighbour if any and setup direction */ LOG_D("search op_start=%p,op_size=0x%lx", operand_start, operand_size); can_expand_fw = can_expand_bw = 0; neighbour = _aspace_bst_search(aspace, operand_start - 1); if (!neighbour || !_compatible(neighbour, operand_size, prop)) { neighbour = _aspace_bst_search(aspace, operand_start + operand_size); if (neighbour && _compatible(neighbour, operand_size, prop)) can_expand_bw = 1; } else can_expand_fw = 1; if (can_expand_fw || can_expand_bw) { /* expand varea at head or tailing */ if (can_expand_fw) operand_start = neighbour->start; operand_size += neighbour->size; LOG_D("expand op_start=%p,op_size=0x%lx", operand_start, operand_size); if (_expand_varea(neighbour, operand_start, operand_size) == RT_EOK) { new_region_at = _merge_surrounding(aspace, neighbour, prop); *pvarea = new_region_at; install_ok = RT_TRUE; } } return install_ok; } static rt_varea_t _varea_create(void *start, rt_size_t size); static int _insert_new_varea(rt_aspace_t aspace, rt_varea_t *pvarea, void *alloc_va, rt_mm_va_hint_t hint) { int err; rt_varea_t varea = *pvarea; if (varea == RT_NULL) { /* no preallocate buffer is provided, then create one */ varea = _varea_create(hint->prefer, hint->map_size); hint->flags &= ~MMF_STATIC_ALLOC; *pvarea = varea; } if (varea) { varea->start = alloc_va; _aspace_bst_insert(aspace, varea); err = RT_EOK; } else { LOG_W("%s: Out of memory", __func__); err = -RT_ENOMEM; } return err; } static inline void _varea_post_install(rt_varea_t varea, rt_aspace_t aspace, rt_size_t attr, rt_size_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { varea->aspace = aspace; varea->attr = attr; varea->mem_obj = mem_obj; varea->flag = flags; varea->offset = offset; if (varea->mem_obj && varea->mem_obj->on_varea_open) varea->mem_obj->on_varea_open(varea); } /** * Install new virtual address region into address space * The pvarea will be set to the varea where new virt memory installed which can * be a newly created or existed one. * * Note: caller must hold the aspace lock */ static int _varea_install(rt_aspace_t aspace, rt_varea_t *pvarea, rt_mm_va_hint_t hint, struct _mapping_property *prop, void **pva) { void *alloc_va; int err = RT_EOK; if (hint->flags & MMF_MAP_FIXED) { alloc_va = hint->prefer; err = _unmap_range_locked(aspace, alloc_va, hint->map_size); if (err != RT_EOK) { /* Note: MAP_FIXED must failed if unable to unmap existing mapping */ LOG_I("%s: unmap range failed in %p with size 0x%lx, error=%d", __func__, alloc_va, hint->map_size, err); } } else { alloc_va = _find_free(aspace, hint->prefer, hint->map_size, hint->limit_start, hint->limit_range_size, hint->flags); if (alloc_va == RT_NULL) err = -RT_ENOSPC; } if (alloc_va != RT_NULL) { /* TODO: fix to private mapping directly */ if (!_try_expand_and_merge_okay(aspace, pvarea, alloc_va, hint, prop)) { err = _insert_new_varea(aspace, pvarea, alloc_va, hint); if (err == RT_EOK) _varea_post_install(*pvarea, aspace, prop->attr, prop->flags, prop->mem_obj, prop->offset); } if (err == RT_EOK) { RT_ASSERT(*pvarea); *pva = alloc_va; } } return err; } /** * restore context modified by varea install */ void _varea_uninstall_locked(rt_varea_t varea) { rt_aspace_t aspace = varea->aspace; if (varea->mem_obj && varea->mem_obj->on_varea_close) varea->mem_obj->on_varea_close(varea); else { rt_hw_mmu_unmap(aspace, varea->start, varea->size); rt_hw_tlb_invalidate_range(aspace, varea->start, varea->size, ARCH_PAGE_SIZE); } _aspace_bst_remove(aspace, varea); } int _mm_aspace_map(rt_aspace_t aspace, rt_varea_t *pvarea, void **addr, rt_size_t length, rt_size_t attr, mm_flag_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { int err = RT_EOK; rt_varea_t varea; struct _mapping_property prop = INIT_PROP(mem_obj, offset, flags, attr); WR_LOCK(aspace); /** * @brief .prefer & .map_size are scratched from varea which setup by caller * .limit_start & .limit_range_size have default to be in range of aspace * .flags is from parameter, and will be fill in varea if install successfully */ struct rt_mm_va_hint hint = {.prefer = *addr, .map_size = length, .limit_start = aspace->start, .limit_range_size = aspace->size, .flags = flags}; if (mem_obj->hint_free) { /* mem object can control mapping range and so by modifing hint */ mem_obj->hint_free(&hint); } /* try to allocate a virtual address region for varea */ err = _varea_install(aspace, pvarea, &hint, &prop, addr); if (err == RT_EOK) { varea = *pvarea; if (MMF_TEST_CNTL(flags, MMF_PREFETCH)) { /* do the MMU & TLB business */ err = _do_prefetch(aspace, varea, varea->start, varea->size); if (err) { LOG_I("%s: failed to prefetch page into varea(%s)", __func__, VAREA_NAME(varea)); /* restore data structure and MMU */ _varea_uninstall_locked(varea); if (!(varea->flag & MMF_STATIC_ALLOC)) rt_free(varea); } } } WR_UNLOCK(aspace); return err; } static rt_varea_t _varea_create(void *start, rt_size_t size) { rt_varea_t varea; varea = (rt_varea_t)rt_malloc(sizeof(struct rt_varea)); if (varea) { varea->start = start; varea->size = size; } return varea; } #define _IS_OVERFLOW(start, length) ((length) > (0ul - (uintptr_t)(start))) #define _IS_OVERSIZE(start, length, limit_s, limit_sz) (((length) + (rt_size_t)((char *)(start) - (char *)(limit_start))) > (limit_size)) static inline int _not_in_range(rt_size_t flags, void *start, rt_size_t length, void *limit_start, rt_size_t limit_size) { /* assuming (base + length) will not overflow except (0) */ int rc = (flags & MMF_MAP_FIXED || start != RT_NULL) ? (_IS_OVERFLOW(start, length) || start < limit_start || _IS_OVERSIZE(start, length, limit_start, limit_size)) : length > limit_size; if (rc) LOG_D("%s: [%p : %p] [%p : %p]", __func__, start, length, limit_start, limit_size); return rc; } static inline int _not_align(void *start, rt_size_t length, rt_size_t mask) { return (start != RT_NULL) && (((uintptr_t)start & mask) || (length & mask)); } /** if the flag is currently supported */ static inline int _not_support(rt_size_t flags) { rt_size_t support_ops = MMF_CREATE(((__MMF_INVALID - 1) << 1) - 1, 1); return flags & ~(support_ops); } int rt_aspace_map(rt_aspace_t aspace, void **addr, rt_size_t length, rt_size_t attr, mm_flag_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { int err; rt_varea_t varea = RT_NULL; RT_DEBUG_SCHEDULER_AVAILABLE(1); if (!aspace || !addr || !mem_obj || length == 0) { err = -RT_EINVAL; LOG_I("%s(%p, %p, %lx, %lx, %lx, %p, %lx): Invalid input", __func__, aspace, addr, length, attr, flags, mem_obj, offset); } else if (_not_in_range(flags, *addr, length, aspace->start, aspace->size)) { err = -RT_EINVAL; LOG_I("%s(addr:%p, len:%lx): out of range", __func__, *addr, length); } else if (_not_support(flags)) { LOG_I("%s: no support flags 0x%lx", __func__, flags); err = -RT_ENOSYS; } else { RT_ASSERT((length & ARCH_PAGE_MASK) == 0); RT_ASSERT(((long)*addr & ARCH_PAGE_MASK) == 0); err = _mm_aspace_map(aspace, &varea, addr, length, attr, flags, mem_obj, offset); } if (err != RT_EOK) { *addr = NULL; } return err; } int rt_aspace_map_static(rt_aspace_t aspace, rt_varea_t varea, void **addr, rt_size_t length, rt_size_t attr, mm_flag_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { int err; if (!aspace || !varea || !addr || !mem_obj || length == 0 || _not_in_range(flags, *addr, length, aspace->start, aspace->size)) { err = -RT_EINVAL; LOG_W("%s: Invalid input", __func__); } else if (_not_support(flags)) { LOG_W("%s: no support flags", __func__); err = -RT_ENOSYS; } else { varea->size = length; varea->start = *addr; flags |= MMF_STATIC_ALLOC; /** * todo: fix if mapping expand, the static varea is not used at all */ err = _mm_aspace_map(aspace, &varea, addr, length, attr, flags, mem_obj, offset); } if (err != RT_EOK) { *addr = NULL; } else { *addr = varea->start; } return err; } int _mm_aspace_map_phy(rt_aspace_t aspace, rt_varea_t varea, rt_mm_va_hint_t hint, rt_size_t attr, rt_size_t pa_off, void **ret_va) { int err; void *vaddr; if (!aspace || !hint || !hint->limit_range_size || !hint->map_size) { LOG_W("%s: Invalid input", __func__); err = -RT_EINVAL; } else if (_not_align(hint->prefer, hint->map_size, ARCH_PAGE_MASK)) { LOG_W("%s: not aligned", __func__); err = -RT_EINVAL; } else if (_not_in_range(hint->flags, hint->limit_start, hint->limit_range_size, aspace->start, aspace->size) || _not_in_range(hint->flags, hint->prefer, hint->map_size, aspace->start, aspace->size)) { LOG_W("%s: not in range", __func__); err = -RT_EINVAL; } else { struct _mapping_property prop = INIT_PROP(0, pa_off, hint->flags, attr); WR_LOCK(aspace); err = _varea_install(aspace, &varea, hint, &prop, &vaddr); if (err == RT_EOK) { err = _do_named_map(aspace, varea, varea->start, varea->size, (rt_size_t)pa_off, attr); if (err != RT_EOK) { _varea_uninstall_locked(varea); } } WR_UNLOCK(aspace); } if (ret_va) { if (err == RT_EOK) *ret_va = vaddr; else *ret_va = RT_NULL; } return err; } int rt_aspace_map_phy(rt_aspace_t aspace, rt_mm_va_hint_t hint, rt_size_t attr, rt_size_t pa_off, void **ret_va) { int err; if (hint) { rt_varea_t varea = _varea_create(hint->prefer, hint->map_size); if (varea) { hint->flags &= ~MMF_STATIC_ALLOC; err = _mm_aspace_map_phy(aspace, varea, hint, attr, pa_off, ret_va); if (err != RT_EOK) { rt_free(varea); } } else { err = -RT_ENOMEM; } } else { err = -RT_EINVAL; } return err; } int rt_aspace_map_phy_static(rt_aspace_t aspace, rt_varea_t varea, rt_mm_va_hint_t hint, rt_size_t attr, rt_size_t pa_off, void **ret_va) { int err; if (varea && hint) { varea->start = hint->prefer; varea->size = hint->map_size; hint->flags |= (MMF_STATIC_ALLOC); LOG_D("%s: start %p size %p phy at %p", __func__, varea->start, varea->size, pa_off << MM_PAGE_SHIFT); err = _mm_aspace_map_phy(aspace, varea, hint, attr, pa_off, ret_va); } else { err = -RT_EINVAL; } return err; } int _aspace_unmap(rt_aspace_t aspace, void *addr) { int error; rt_varea_t varea; WR_LOCK(aspace); varea = _aspace_bst_search(aspace, addr); if (varea == RT_NULL) { LOG_D("%s: No such entry found at %p\n", __func__, addr); error = -RT_ENOENT; } else { _varea_uninstall_locked(varea); if (!(varea->flag & MMF_STATIC_ALLOC)) { rt_free(varea); } error = RT_EOK; } WR_UNLOCK(aspace); return error; } int rt_aspace_unmap(rt_aspace_t aspace, void *addr) { int error; if (!aspace) { LOG_I("%s: Invalid input", __func__); error = -RT_EINVAL; } else if (_not_in_range(MMF_MAP_FIXED, addr, 1, aspace->start, aspace->size)) { LOG_I("%s: %lx not in range of aspace[%lx:%lx]", __func__, addr, aspace->start, (char *)aspace->start + aspace->size); error = -RT_EINVAL; } else { error = _aspace_unmap(aspace, addr); } return error; } /** * modify the property of existed varea by shrink its size. Mem_obj is * notified to released the resource. */ static rt_err_t _shrink_varea(rt_varea_t varea, void *new_va, rt_size_t size) { rt_err_t error; rt_aspace_t aspace; void *old_va; if (varea->mem_obj && varea->mem_obj->on_varea_shrink) error = varea->mem_obj->on_varea_shrink(varea, new_va, size); else error = -RT_EPERM; if (error == RT_EOK) { aspace = varea->aspace; old_va = varea->start; varea->size = size; if (old_va != new_va) { varea->start = new_va; varea->offset += ((long)new_va - (long)old_va) >> MM_PAGE_SHIFT; _aspace_bst_remove(aspace, varea); _aspace_bst_insert(aspace, varea); } } return error; } static rt_err_t _split_varea(rt_varea_t existed, char *ex_end, char *unmap_start, char *unmap_end, rt_size_t unmap_len) { int error; size_t rela_offset; rt_varea_t subset; char *subset_start; size_t subset_size; if (existed->mem_obj && existed->mem_obj->on_varea_split) { subset_start = unmap_end; subset_size = ex_end - subset_start; subset = _varea_create(subset_start, subset_size); if (subset) { rela_offset = MM_PA_TO_OFF(subset_start - (char *)existed->start); subset->aspace = existed->aspace; subset->attr = existed->attr; subset->mem_obj = existed->mem_obj; subset->flag = existed->flag & ~MMF_STATIC_ALLOC; subset->offset = existed->offset + rela_offset; error = existed->mem_obj->on_varea_split(existed, unmap_start, unmap_len, subset); if (error == RT_EOK) { existed->size = unmap_start - (char *)existed->start; _aspace_bst_insert(existed->aspace, subset); } if (error != RT_EOK) rt_free(subset); } else error = -RT_ENOMEM; } else error = -RT_EPERM; return error; } /* remove overlapped pages from varea */ static int _remove_overlapped_varea(rt_varea_t existed, char *unmap_start, rt_size_t unmap_len) { int error; char *ex_start = existed->start; char *ex_end = ex_start + existed->size; char *unmap_end = unmap_start + unmap_len; if (ex_start < unmap_start) { if (ex_end > unmap_end) error = _split_varea(existed, ex_end, unmap_start, unmap_end, unmap_len); else error = _shrink_varea(existed, ex_start, unmap_start - ex_start); } else if (ex_end > unmap_end) error = _shrink_varea(existed, unmap_end, ex_end - unmap_end); else { _varea_uninstall_locked(existed); if (VAREA_NOT_STATIC(existed)) { rt_free(existed); } error = RT_EOK; } return error; } static int _unmap_range_locked(rt_aspace_t aspace, void *addr, size_t length) { int error = RT_EOK; rt_varea_t existed; struct _mm_range unmap_range; unmap_range.start = addr; unmap_range.end = addr + length - 1; existed = _aspace_bst_search_overlap(aspace, unmap_range); while (existed) { error = _remove_overlapped_varea(existed, addr, length); if (error == RT_EOK) existed = _aspace_bst_search_overlap(aspace, unmap_range); else break; } return error; } int rt_aspace_unmap_range(rt_aspace_t aspace, void *addr, size_t length) { int error; if (!aspace) { LOG_I("%s: Invalid input", __func__); error = -RT_EINVAL; } else if (_not_in_range(MMF_MAP_FIXED, addr, length, aspace->start, aspace->size)) { LOG_I("%s: %lx not in range of aspace[%lx:%lx]", __func__, addr, aspace->start, (char *)aspace->start + aspace->size); error = -RT_EINVAL; } else if (!ALIGNED(addr)) { LOG_I("%s(addr=%p): Unaligned address", __func__, addr); error = -RT_EINVAL; } else { /** * Brief: re-arrange the address space to remove existing pages mapping * in [unmap_start, unmap_start + unmap_len) */ length = RT_ALIGN(length, ARCH_PAGE_SIZE); WR_LOCK(aspace); error = _unmap_range_locked(aspace, addr, length); WR_UNLOCK(aspace); } return error; } void *rt_aspace_mremap_range(rt_aspace_t aspace, void *old_address, size_t old_size, size_t new_size, int flags, void *new_address) { void *ret = RT_NULL; if (!aspace) { LOG_I("%s: Invalid input", __func__); } else if (_not_in_range(MMF_MAP_FIXED, old_address, old_size, aspace->start, aspace->size)) { LOG_I("%s: %lx not in range of aspace[%lx:%lx]", __func__, old_address, aspace->start, (char *)aspace->start + aspace->size); } else if (!ALIGNED(old_address)) { LOG_I("%s(old_address=%p): Unaligned address", __func__, old_address); } else { /** * Brief: re-arrange the address space to remove existing pages mapping * in [unmap_start, unmap_start + unmap_len) */ old_size = RT_ALIGN(old_size, ARCH_PAGE_SIZE); WR_LOCK(aspace); { rt_varea_t existed; struct _mm_range unmap_range; unmap_range.start = old_address; unmap_range.end = old_address + old_size - 1; existed = _aspace_bst_search_overlap(aspace, unmap_range); if (existed && existed->mem_obj && existed->mem_obj->on_varea_mremap) { ret = existed->mem_obj->on_varea_mremap(existed, new_size, flags, new_address); } } WR_UNLOCK(aspace); if (ret) { int error = rt_aspace_unmap_range(aspace, old_address, old_size); if (error != RT_EOK) { LOG_I("%s: unmap old failed, addr %p size %p", __func__, old_address, old_size); } } } return ret; } static inline void *_lower(void *a, void *b) { return a < b ? a : b; } static inline void *_align(void *va, rt_ubase_t align_mask) { return (void *)((rt_ubase_t)((char *)va + ~align_mask) & align_mask); } static void *_ascending_search(rt_varea_t varea, rt_size_t req_size, rt_ubase_t align_mask, struct _mm_range limit) { void *ret = RT_NULL; while (varea && varea->start < limit.end) { char *candidate = (char *)varea->start + varea->size; candidate = _align(candidate, align_mask); if (candidate > (char *)limit.end || (char *)limit.end - candidate + 1 < req_size) break; rt_varea_t nx_va = ASPACE_VAREA_NEXT(varea); if (nx_va) { rt_size_t gap_size = (char *)_lower(limit.end, (char *)nx_va->start - 1) - candidate + 1; if (gap_size >= req_size) { ret = candidate; break; } } else { ret = candidate; } varea = nx_va; } return ret; } /** find suitable place in [limit_start, limit_end] */ static void *_find_head_and_asc_search(rt_aspace_t aspace, rt_size_t req_size, rt_ubase_t align_mask, struct _mm_range limit) { void *va = RT_NULL; rt_varea_t varea = _aspace_bst_search_exceed(aspace, limit.start); if (varea) { char *candidate = _align(limit.start, align_mask); rt_size_t gap_size = (char *)varea->start - candidate; if (gap_size >= req_size) { rt_varea_t former = _aspace_bst_search(aspace, limit.start); if (former) { candidate = _align((char *)former->start + former->size, align_mask); gap_size = (char *)varea->start - candidate; if (gap_size >= req_size) va = candidate; else va = _ascending_search(varea, req_size, align_mask, limit); } else { va = candidate; } } else { va = _ascending_search(varea, req_size, align_mask, limit); } } else { char *candidate; rt_size_t gap_size; candidate = limit.start; candidate = _align(candidate, align_mask); gap_size = (char *)limit.end - candidate + 1; if (gap_size >= req_size) va = candidate; } return va; } static void *_find_free(rt_aspace_t aspace, void *prefer, rt_size_t req_size, void *limit_start, rt_size_t limit_size, mm_flag_t flags) { rt_varea_t varea = NULL; void *va = RT_NULL; struct _mm_range limit = {limit_start, (char *)limit_start + limit_size - 1}; rt_ubase_t align_mask = ~0ul; if (flags & MMF_REQUEST_ALIGN) { align_mask = ~((1 << MMF_GET_ALIGN(flags)) - 1); } if (prefer != RT_NULL) { /* if prefer and free, just return the prefer region */ prefer = _align(prefer, align_mask); struct _mm_range range = {prefer, (char *)prefer + req_size - 1}; varea = _aspace_bst_search_overlap(aspace, range); if (!varea) { va = prefer; } else if (flags & MMF_MAP_FIXED) { /* OVERLAP */ } else { /* search from `varea` in ascending order */ va = _ascending_search(varea, req_size, align_mask, limit); if (va == RT_NULL) { /* rewind to first range */ limit.end = (char *)varea->start - 1; va = _find_head_and_asc_search(aspace, req_size, align_mask, limit); } } } else { va = _find_head_and_asc_search(aspace, req_size, align_mask, limit); } return va; } int rt_aspace_load_page(rt_aspace_t aspace, void *addr, rt_size_t npage) { int err = RT_EOK; rt_varea_t varea; char *end = (char *)addr + (npage << ARCH_PAGE_SHIFT); WR_LOCK(aspace); varea = _aspace_bst_search(aspace, addr); WR_UNLOCK(aspace); if (!varea) { LOG_W("%s: varea not exist(addr=%p)", __func__, addr); err = -RT_ENOENT; } else if ((char *)addr >= end || (rt_size_t)addr & ARCH_PAGE_MASK || _not_in_range(MMF_MAP_FIXED, addr, npage << ARCH_PAGE_SHIFT, varea->start, varea->size)) { LOG_W("%s: Unaligned parameter or out of range", __func__); err = -RT_EINVAL; } else { err = _do_prefetch(aspace, varea, addr, npage << ARCH_PAGE_SHIFT); } return err; } int rt_varea_map_page(rt_varea_t varea, void *vaddr, void *page) { int err = RT_EOK; void *page_pa = rt_kmem_v2p(page); if (!varea || !vaddr || !page) { LOG_W("%s(%p,%p,%p): invalid input", __func__, varea, vaddr, page); err = -RT_EINVAL; } else if (page_pa == ARCH_MAP_FAILED) { LOG_W("%s: page is not in kernel space", __func__); err = -RT_ERROR; } else if (_not_in_range(MMF_MAP_FIXED, vaddr, ARCH_PAGE_SIZE, varea->start, varea->size)) { LOG_W("%s(%p,%lx): not in range of varea(%p,%lx)", __func__, vaddr, ARCH_PAGE_SIZE, varea->start, varea->size); err = -RT_EINVAL; } else { err = _do_named_map( varea->aspace, varea, vaddr, ARCH_PAGE_SIZE, MM_PA_TO_OFF(page_pa), varea->attr ); } return err; } int rt_varea_unmap_page(rt_varea_t varea, void *vaddr) { void *va_aligned = (void *)RT_ALIGN_DOWN((rt_base_t)vaddr, ARCH_PAGE_SIZE); return rt_varea_unmap_range(varea, va_aligned, ARCH_PAGE_SIZE); } /** * @note Caller should take care of synchronization of its varea among all the map/unmap operation */ int rt_varea_map_range(rt_varea_t varea, void *vaddr, void *paddr, rt_size_t length) { int err; if (!varea || !vaddr || !paddr || !length || !ALIGNED(vaddr) || !ALIGNED(paddr) || !(ALIGNED(length))) { LOG_W("%s(%p,%p,%p,%lx): invalid input", __func__, varea, vaddr, paddr, length); err = -RT_EINVAL; } else if (_not_in_range(MMF_MAP_FIXED, vaddr, length, varea->start, varea->size)) { LOG_W("%s(%p,%lx): not in range of varea(%p,%lx)", __func__, vaddr, length, varea->start, varea->size); err = -RT_EINVAL; } else { err = _do_named_map( varea->aspace, varea, vaddr, length, MM_PA_TO_OFF(paddr), varea->attr ); } return err; } /** * @note Caller should take care of synchronization of its varea among all the map/unmap operation */ int rt_varea_unmap_range(rt_varea_t varea, void *vaddr, rt_size_t length) { int err; rt_base_t va_align; if (!varea || !vaddr || !length) { LOG_W("%s(%p,%p,%lx): invalid input", __func__, varea, vaddr, length); err = -RT_EINVAL; } else if (_not_in_range(MMF_MAP_FIXED, vaddr, length, varea->start, varea->size)) { LOG_W("%s(%p,%lx): not in range of varea(%p,%lx)", __func__, vaddr, length, varea->start, varea->size); err = -RT_EINVAL; } else { va_align = RT_ALIGN_DOWN((rt_base_t)vaddr, ARCH_PAGE_SIZE); rt_hw_mmu_unmap(varea->aspace, (void *)va_align, length); rt_hw_tlb_invalidate_range(varea->aspace, (void *)va_align, length, ARCH_PAGE_SIZE); err = RT_EOK; } return err; } int rt_aspace_offload_page(rt_aspace_t aspace, void *addr, rt_size_t npage) { return -RT_ENOSYS; } int rt_aspace_control(rt_aspace_t aspace, void *addr, enum rt_mmu_cntl cmd) { int err; rt_varea_t varea; WR_LOCK(aspace); varea = _aspace_bst_search(aspace, addr); WR_UNLOCK(aspace); if (varea) { err = rt_hw_mmu_control(aspace, varea->start, varea->size, cmd); if (err == RT_EOK) { rt_hw_tlb_invalidate_range(aspace, varea->start, varea->size, ARCH_PAGE_SIZE); } } else { err = -RT_ENOENT; } return err; } int rt_aspace_traversal(rt_aspace_t aspace, int (*fn)(rt_varea_t varea, void *arg), void *arg) { rt_varea_t varea; rt_varea_t next; WR_LOCK(aspace); varea = ASPACE_VAREA_FIRST(aspace); while (varea) { next = ASPACE_VAREA_NEXT(varea); fn(varea, arg); varea = next; } WR_UNLOCK(aspace); return 0; } static int _dump(rt_varea_t varea, void *arg) { if (varea->mem_obj && varea->mem_obj->get_name) { rt_kprintf("[%p - %p] %s\n", varea->start, (char *)varea->start + varea->size, varea->mem_obj->get_name(varea)); } else { rt_kprintf("[%p - %p] phy-map\n", varea->start, (char *)varea->start + varea->size); rt_kprintf("\t\\_ paddr = %p\n", varea->offset << MM_PAGE_SHIFT); } return 0; } void rt_aspace_print_all(rt_aspace_t aspace) { rt_aspace_traversal(aspace, _dump, NULL); } static int _count_vsz(rt_varea_t varea, void *arg) { rt_base_t *pvsz = arg; RT_ASSERT(varea); *pvsz = *pvsz + varea->size; return 0; } rt_base_t rt_aspace_count_vsz(rt_aspace_t aspace) { rt_base_t vsz = 0; rt_aspace_traversal(aspace, _count_vsz, &vsz); return vsz; } static int _dup_varea(rt_varea_t src_varea, void *arg) { int err; rt_aspace_t dst = arg; rt_aspace_t src = src_varea->aspace; void *pa = RT_NULL; void *va = RT_NULL; rt_mem_obj_t mem_obj = src_varea->mem_obj; if (!mem_obj) { /* duplicate a physical mapping */ pa = rt_hw_mmu_v2p(src, (void *)src_varea->start); RT_ASSERT(pa != ARCH_MAP_FAILED); struct rt_mm_va_hint hint = {.flags = src_varea->flag, .limit_range_size = dst->size, .limit_start = dst->start, .prefer = src_varea->start, .map_size = src_varea->size}; err = rt_aspace_map_phy(dst, &hint, src_varea->attr, MM_PA_TO_OFF(pa), &va); if (err != RT_EOK) { LOG_W("%s: aspace map failed at %p with size %p", __func__, src_varea->start, src_varea->size); } } else { /* duplicate a mem_obj backing mapping */ rt_base_t flags = src_varea->flag | MMF_MAP_FIXED; flags &= ~MMF_STATIC_ALLOC; flags &= ~MMF_PREFETCH; va = src_varea->start; err = rt_aspace_map(dst, &va, src_varea->size, src_varea->attr, flags, mem_obj, src_varea->offset); if (err != RT_EOK) { LOG_W("%s: aspace map failed at %p with size %p", __func__, src_varea->start, src_varea->size); } } if (va != (void *)src_varea->start) { return -1; } return 0; } struct _compare_param { rt_aspace_t dst; int rc; }; rt_err_t rt_aspace_duplicate_locked(rt_aspace_t src, rt_aspace_t dst) { return rt_aspace_traversal(src, _dup_varea, dst); } rt_inline int _varea_same(rt_varea_t a, rt_varea_t b) { return a->attr == b->attr && a->flag == b->flag && a->mem_obj == b->mem_obj; } rt_inline void _dump_varea(rt_varea_t varea) { LOG_W("%s(attr=0x%lx, flags=0x%lx, start=0x%lx, size=0x%lx, mem_obj=%p)", VAREA_NAME(varea), varea->attr, varea->flag, varea->start, varea->size, varea->mem_obj); } static int _compare_varea(rt_varea_t src_varea, void *arg) { struct _compare_param *param = arg; rt_varea_t dst_varea; rt_aspace_t dst = param->dst; rt_aspace_t src = src_varea->aspace; dst_varea = _aspace_bst_search(dst, src_varea->start); if (dst_varea) { char *buf1 = rt_pages_alloc_ext(0, PAGE_ANY_AVAILABLE); char *buf2 = rt_pages_alloc_ext(0, PAGE_ANY_AVAILABLE); char *vend = src_varea->start + src_varea->size; for (char *i = src_varea->start; i < vend; i += ARCH_PAGE_SIZE) { int rc; rt_aspace_page_get(src, i, buf1); rt_aspace_page_get(dst, i, buf2); rc = memcmp(buf1, buf2, ARCH_PAGE_SIZE); if (rc) { if (param->rc == 0) param->rc = rc; LOG_E("%s(a_varea=%s, b_varea=%s)", __func__, VAREA_NAME(src_varea), VAREA_NAME(dst_varea)); _dump_varea(src_varea); _dump_varea(dst_varea); RT_ASSERT(0); } } rt_pages_free(buf1, 0); rt_pages_free(buf2, 0); } else { param->rc = -RT_ENOENT; } return 0; } rt_err_t rt_aspace_compare(rt_aspace_t src, rt_aspace_t dst) { struct _compare_param param = {.rc = 0, .dst = dst}; rt_aspace_traversal(src, _compare_varea, ¶m); return param.rc; } /* dst are page aligned */ rt_inline rt_err_t _page_put(rt_varea_t varea, void *page_va, void *buffer) { struct rt_aspace_io_msg iomsg; rt_err_t rc; rt_mm_io_msg_init(&iomsg, VAREA_VA_TO_OFFSET(varea, page_va), page_va, buffer); varea->mem_obj->page_write(varea, &iomsg); if (iomsg.response.status == MM_FAULT_STATUS_UNRECOVERABLE) rc = -RT_ERROR; else rc = RT_EOK; return rc; } /* dst are page aligned */ rt_inline rt_err_t _page_get(rt_varea_t varea, void *page_va, void *buffer) { struct rt_aspace_io_msg iomsg; rt_err_t rc; rt_mm_io_msg_init(&iomsg, VAREA_VA_TO_OFFSET(varea, page_va), page_va, buffer); varea->mem_obj->page_read(varea, &iomsg); if (iomsg.response.status == MM_FAULT_STATUS_UNRECOVERABLE) rc = -RT_ERROR; else rc = RT_EOK; return rc; } #ifdef RT_USING_SMART #include "lwp.h" rt_inline rt_aspace_t _current_uspace(void) { rt_lwp_t this_proc = lwp_self(); return this_proc ? this_proc->aspace : RT_NULL; } #else rt_inline rt_aspace_t _current_uspace(void) { return RT_NULL; } #endif rt_err_t rt_aspace_page_get_phy(rt_aspace_t aspace, void *page_va, void *buffer) { rt_err_t rc = -RT_ERROR; char *frame_ka = rt_hw_mmu_v2p(aspace, page_va); if (frame_ka != ARCH_MAP_FAILED) { frame_ka = rt_kmem_p2v(frame_ka); if (frame_ka) { rt_memcpy(buffer, frame_ka, ARCH_PAGE_SIZE); rc = RT_EOK; } else if (aspace == _current_uspace() || aspace == &rt_kernel_space) { /* direct IO */ rt_memcpy(buffer, page_va, ARCH_PAGE_SIZE); rc = RT_EOK; } else { /* user memory region remap ? */ LOG_W("%s(aspace=0x%lx,va=%p): Operation not support", __func__, aspace, page_va); rc = -RT_ENOSYS; } } else { LOG_W("%s(aspace=0x%lx,va=%p): PTE not existed", __func__, aspace, page_va); rc = -RT_ENOENT; } return rc; } rt_err_t rt_aspace_page_put_phy(rt_aspace_t aspace, void *page_va, void *buffer) { rt_err_t rc = -RT_ERROR; char *frame_ka = rt_hw_mmu_v2p(aspace, page_va); if (frame_ka != ARCH_MAP_FAILED) { frame_ka = rt_kmem_p2v(frame_ka); if (frame_ka) { rt_memcpy(frame_ka, buffer, ARCH_PAGE_SIZE); rc = RT_EOK; } else if (aspace == _current_uspace() || aspace == &rt_kernel_space) { /* direct IO */ rt_memcpy(page_va, buffer, ARCH_PAGE_SIZE); rc = RT_EOK; } else { /* user memory region remap ? */ LOG_W("%s(aspace=0x%lx,va=%p): Operation not support", __func__, aspace, page_va); rc = -RT_ENOSYS; } } else { LOG_W("%s(aspace=0x%lx,va=%p): PTE not existed", __func__, aspace, page_va); rc = -RT_ENOENT; } return rc; } rt_err_t rt_aspace_page_put(rt_aspace_t aspace, void *page_va, void *buffer) { rt_err_t rc = -RT_ERROR; rt_varea_t varea; RT_ASSERT(aspace); RD_LOCK(aspace); varea = _aspace_bst_search(aspace, page_va); if (varea && ALIGNED(page_va)) { if (varea->mem_obj) { if (varea->mem_obj->page_write) { if (rt_varea_is_private_locked(varea)) { RDWR_LOCK(aspace); struct rt_aspace_fault_msg msg; msg.fault_op = MM_FAULT_OP_WRITE; msg.fault_type = MM_FAULT_TYPE_GENERIC_MMU; msg.fault_vaddr = page_va; rc = rt_varea_fix_private_locked(varea, rt_hw_mmu_v2p(aspace, page_va), &msg, RT_TRUE); RDWR_UNLOCK(aspace); if (rc == MM_FAULT_FIXABLE_TRUE) { varea = _aspace_bst_search(aspace, page_va); rc = _page_put(varea, page_va, buffer); } else rc = -RT_ERROR; } else rc = _page_put(varea, page_va, buffer); } else { rc = -RT_EINVAL; LOG_I("%s: Operation not allowed", __func__); } } else { rc = rt_aspace_page_put_phy(aspace, page_va, buffer); } } else rc = -RT_EINVAL; RD_UNLOCK(aspace); return rc; } rt_err_t rt_aspace_page_get(rt_aspace_t aspace, void *page_va, void *buffer) { rt_err_t rc = -RT_ERROR; rt_varea_t varea; /* TODO: cache the last search item */ RT_ASSERT(aspace); RD_LOCK(aspace); varea = _aspace_bst_search(aspace, page_va); if (varea && ALIGNED(page_va)) { if (varea->mem_obj) { if (varea->mem_obj->page_read) { rc = _page_get(varea, page_va, buffer); } else { LOG_I("%s: Operation not allowed", __func__); } } else { rc = rt_aspace_page_get_phy(aspace, page_va, buffer); } } else { rc = -RT_EINVAL; LOG_D("%s(va=%p,varea=0x%lx): Invalid address", __func__, page_va, varea); } RD_UNLOCK(aspace); return rc; } rt_varea_t rt_aspace_query(rt_aspace_t aspace, void *vaddr) { return _aspace_bst_search(aspace, vaddr); }